Nvidia's Strategy; What Next

But there are many difficulties on both parts, and both are far from sure things.

Nvidia's chosen field is one of the fastest moving and most competitive on the planet.  It's successful strategy of 1998-2008 does not ensure enduring success.  In particular, by 2009 the waves of change Nvidia had exploited in its rise were waning.   

Strategically, Nvidia is presently engaged in a two-pronged pincer movement.  One initiative has been opening up access to the computing power of its graphics chips.  Each graphics processor contains hundreds of separate floating-point processors.  Its new hardware supporting this concept is called Telsa and delivers desktop supercomputing.  In November, 2010, Chinese researchers announced the world's tastes supercomputer-powered by Nvidia Tesla graphics chips.

The second arm of the pincer is Tegra--a complete system on a chip. This tactic is called "disruption from below" and aims to upset the Intel-AMD-Windows hegemony by building on a much simpler, more efficient platform.  The chip is aimed at the makers of smart phones, net books, and game consoles.   

This pincer movement offers two ways to succeed and makes trouble for competitors.  But there are many difficulties on both parts, and both are far from sure things.

ACTION POINT: Strategy is an ongoing consideration for every business.

Strategy at Nvidia; The Competition

The deeper reality was that Nvidia's carefully crafted fast-release cycle induced 3dfx's less coordinated responses. 

For Nvidia's strategy in 3-D graphics to work, other firms would have to fall by the wayside, not being able to keep up.  To a large extent, that is exactly what happened.

Whenever a company succeeds greatly there is a complementary story of impeded competitive response.  Sometimes the impediment is the innovator's patent or similar protection, but more often it is an unwillingness or inability to replicate the innovator's policies.  Nvidia bet that it could rapidly charge up the graphics pipeline, but it also be that key rivals would not be able to replicate its fast release cycle.  Rival 3dfx followed spurious advice from Wall Street and the marketing instincts of a new CEO, setting out after the mass market.  Spreading its resources too thin, it attempted to compensate by stretching the goals for its next high-performance chip beyond the competencies of its development process.  

3dfx closed its doors in the last months of 2000, selling its patents, brands, and inventory to Nvidia where many of its talented engineers ended up working.   A surface reading of history makes it look like 3dfx did itself in with too many changes of direction.  The deeper reality was that Nvidia's carefully crafted fast-release cycle induced 3dfx's less coordinated responses. 

ACTION POINT: As Hannibal did to Rome at Cannae, Nvidia enticed its rival into overreaching.

Strategy at Nvidia: Cohesive Action

The benefit of a faster cycle is that the product will be best in class more often.

The first step in executing the guiding policy of a faster cycle time  was the establishment of three separate development teams.  Each would work  to an eighteen-month start-to-market cycle.  With overlapping schedules, the three teams would deliver a new product every six months.  Accordingly, the second set of policies were meant to substantially reduce the delays and uncertainty in the development process. 

A serious source of possible delays was a design error.  To address this source of possible delays, Nvidia invested heavily in simulation and emulation techniques and organized its chip design process around these methods.   Nvidia also took control of the creation and management of drivers for its chips, developing a unified driver architecture.  This approach would greatly simplify things for users because they would not have to be concerned with matching drivers to chips.   To speed up driver development, the company made a significant investment in emulation facilities.  These were complex hardware "mock-ups" of new chips that allowed driver development to begin four to six months before the first true chips appeared.  Implementing the new strategy, Nvidia invested its remaining cash in emulation equipment and in engineering a new chip.   The benefit of a faster cycle is that the product will be best in class more often.  Compared to a competitor working on an eighteen-month cycle, Nvidia's six-month cycle would mean that its chip would be the better product about 83 percent of the time.

   

ACTION POINT: Act and invest in support of the guiding policy.

Strategy at Nvidia; Guiding Policy

...the Nvidia team designed a set of cohesive policies and actions to turn their guiding policy into reality.

Progress in the semiconductor industry was based on reducing the size of a transistor.  Smaller transistors were faster and consumed less power.  The whole semiconductor industry coordinated around achieving a higher level of integration, based on smaller transistors, about every eighteen months.  This rate of progress was called Moore's law.  No one could jump much ahead of this pace because all the technologies, from photo lithography to optical design to metal deposition to testing, had to advance in lockstep.   The industry called this patter of collective advance the "road map."

Nvidia's top management and its technology board sketched out a different road map, one that would push 3-D graphics performance at a much faster rate than Moore's law predicted.  There were two factors crucial to making this happen.  First, they anticipated a large jump in performance from putting more and more of the graphics pipeline onto a single chip.  Second, most manufacturers, such as Intel, didn't keep cramming the maximum number of transistors possible onto a chip--they took advantage of increases in density to make more chis per wafer, thereby cutting the cost of chip.  Nvidia, by contrast, planned to use that extra density to add more parallel processors, boosting performance-a consequence of using the Utah-SGI "triangle bus" idea.

On the demand side, management judged that the market would buy virtually all the graphics processing power that could be provided.  There was not shape call for word processors or spreadsheets but ran one hundred times faster, but there was a very sharp demand for faster and more realistic graphics.  CEO Jen-Hsun Huang believed that Nvidia could construct an advantage by breaking the industry's eighteen month cycle.   He reasoned that since it looked possible to advance graphics power three times faster than CPU poet, Nvidia could deliver a substantial upgrade in graphics power every six months instead of every eighteen months.  

This is the point where a bad strategist would have wrapped the concept of a faster development cycle in slogans about sped, power, and growth, and then sought to cash in by taking the company public.  Instead, the Nvidia team designed a set of cohesive policies and actions to turn their guiding policy into reality.

ACTION POINT: Sometimes a guiding policy will go against the grain of the industry standard.

Creating a True Whole

Create a true whole greater than the sum of its parts.

A manager has the task of creating a true whole that is larger than the sum of its parts.  One analogy is the task of the conductor of a symphony orchestra, through whose effort, vision, and leadership individual instrumental parts become the living whole of a musical performance.  But the conductor has the composer’s score; he is only interpreter.  The manager is both composer and conductor.

The task of creating a genuine whole also requires that the manager, in every one of her acts, consider simultaneously the performance and results of the enterprise as a whole and the diverse activities needed to achieve synchronized performance.  It is here, perhaps,that the comparison with the orchestra conductor fits best. A conductor must always hear both the whole orchestra and, say, the second oboe.  Similarly, a manager must always consider both the overall performance of the enterprise and, say, the market-research activity needed.  By raising the performance of the whole, she creates scope and challenge for market research.  By improving the performance of market research, she makes possible better overall business results. 

The manager must simultaneously ask two double-barreled questions: “What better business performance is needed and what does this require of what activities?”  AND “What better performances are the activities capable of and what improvement in business results will they make possible?”

ACTION POINT: Have you composed your symphony?